A Cross-Sectional Study of Psychosocial Work Environment and Stress in the Danish Symphony Orchestras

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/51731255

A cross-sectional study of psychosocial work environment and stress in the Danish symphony orchestras

Article in International Archives of Occupational and Environmental Health · October 2011 DOI: 10.1007/s00420-011-0710-z · Source: PubMed

CITATIONS READS 35 401

3 authors, including:

Helene M Paarup Jesper Baelum University of Southern Denmark Odense University Hospital

21 PUBLICATIONS 282 CITATIONS 96 PUBLICATIONS 1,911 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Musicians' Health - Denmark View project

Reports on Occupational Lung Disease View project

All content following this page was uploaded by Helene M Paarup on 25 January 2015.

The user has requested enhancement of the downloaded file. Int Arch Occup Environ Health DOI 10.1007/s00420-011-0710-z

ORIGINAL ARTICLE

A cross-sectional study of psychosocial work environment and stress in the Danish symphony orchestras

Gitte Juel Holst · Helene M. Paarup · Jesper Baelum

Received: 20 February 2011 / Accepted: 4 October 2011 © Springer-Verlag 2011

Abstract associated with increasing stress symptoms. Compared to Purpose To investigate psychosocial work environment the general workforce independently of gender, Danish and stress in Danish symphony orchestra musicians. symphony orchestra musicians reported higher emotional Methods This was a cross-sectional questionnaire survey demands, lower inXuence, lower social support, lower of psychosocial work factors and stress symptoms among sense of community, and lower job satisfaction. However, 441 musicians in six Danish symphony orchestras. The the musicians reported a higher commitment to the work- response rate was 78% (n = 342). The questions were from place. COPSOQ (Copenhagen Psychosocial Questionnaire). Conclusions The Wndings indicate a more demanding Mean values of 19 COPSOQ-scales were compared by gen- psychosocial work environment exposure among sym- der and instrument group. The results for the musicians phony orchestra musicians than among Danish workers in were compared with results for the general Danish work general. Critical results are the relatively high work force (COPSOQ database). demands, low inXuence, and low social support, females Results Female musicians reported higher work demands being of higher risk than males. and higher stress symptoms than their male colleagues. Between instrument groups, 2nd violinists seemed to be of Keywords Psychosocial work environment exposure · particular risk compared with the other instrumental groups Stress symptoms · Symphony orchestra musicians in aspects of work pace, work organization, and content, whereas 1st violinists perceived higher emotional stress compared with 2nd violinists. The musicians’ experience of Introduction increased work demands as well as deteriorated, work organization and job content, interpersonal relations and The hierarchical structure of the symphony orchestra leaves leadership, and work-individual interface was signiWcantly limited inXuence for the individual musician, and the psychosocial work environment of symphony orchestra musicians has been characterized by rigorous demands for high levels of technical skills, strict discipline, and an orchestral G. J. Holst · H. M. Paarup · J. Baelum Department of Occupational and Environmental Medicine, performance depending on high collaborative skills (Ost- Odense University Hospital, Odense, Denmark wald et al. 1994; Fetter 1993; Sternbach 1993; Piperek 1981). Thus, symphony orchestra musicians are a highly & G. J. Holst ( ) selected group. Despite these remarkable working condi- Department of Environmental and Occupational Health, Aarhus University, Ole Worms Allé 6, 8000 Aarhus C, Denmark tions, only few epidemiological studies have emphasized e-mail: [email protected] on the psychosocial work environment and stress of symphony orchestra musicians, and the results of the studies H. M. Paarup · J. Baelum have been conXicting. A Danish study showed that only 1% Research Unit of Occupational and Environmental Medicine, Clinical Institute, University of Southern Denmark, of 145 symphony orchestra musicians reported they often Odense, Denmark experienced stress (Laitinen and Poulsen 2008). In contrast, 123 Int Arch Occup Environ Health an American study of 2,212 musicians from 47 symphony Having the resources to fulWll the speciWc work condi- orchestras showed that 23% of the musicians reported more tions, the symphony orchestra musicians represent a perceived stress than colleague musicians (Middlestadt and selected group, but a paradox seems to exist; on one hand, Fishbein 1988). Similarly, among 154 musicians, 21% studies have reported that symphony orchestra musicians experienced moderate stress, 6% severe, and 4% extreme experience strenuous work conditions and stress; on the stress (Salmon et al. 1995). other hand, symphony orchestra musicians also report high Previous studies have identiWed string and woodwind levels of work satisfaction (Kivimäki and Jokinen 1994; players to report more stress symptoms than other instru- Liljeholm Johansson 1996). Many studies on psychosocial ment groups (Middlestadt and Fishbein 1988; Parasur- work environment and stress among diVerent occupational aman and Purohit 2000; Liljeholm Johansson and groups have been performed over the last decades, in Den- Theorell 2003). However, most studies have not distin- mark as well as in many other countries. However, sym- guished between string players, which should be of partic- phony orchestra musicians have often been neglected in ular interest since 1st violinists seem to be more exposed these studies leaving an uncertainty about their psychoso- due to a high work load, high technical demands, being cial work environment exposure and thereby how this pro- responsible for solos, and being more visible during the fession is characterized compared with other professions. entire performance, while the 2nd violins usually are a The overall objective of the present study was to investi- subordinate to the 1st violins. Thus, the 2nd violinists gate psychosocial work environment and stress in Danish seem more withdrawn as support for the remaining symphony orchestra musicians emphasizing on three ques- orchestra which may result in perceiving the psychosocial tions: (1) Does the mean level of work stress diVer between work environment less strenuous. Moreover, the inXuence diVerent groups within the orchestras?; (2) Is adverse psy- of gender on stress symptoms seems to be conXicting. chosocial work environment positively associated with Liljeholm and Theorell found a signiWcant diVerence stress symptoms?; (3) Do mean levels of works stress and between genders regarding female musicians reported a psychosocial work environment of symphony orchestra more adverse work environment compared with male musicians diVer from those of the general work force? This musicians (Liljeholm Johansson and Theorell 2003). In knowledge may be helpful for health professionals to get a contrast to this, Hamilton et al. found that male orchestra better understanding of the musicians work conditions and musicians reported higher mental, physical, and interper- allow health professionals to advise and treat symphony sonal strain than female musicians (Hamilton et al. 1995). orchestra musicians appropriately. A third study found that the perception of stress was unre- lated to instrumental group and gender in 93 musicians (Kivimäki and Jokinen 1994). Methods Occupational stress is considered a signiWcant work- related health problem (European Commission 2002). Participants Therefore, it is of great preventive interest to get an understanding of how musicians experience their psychosocial Study population work environment and to what extent stress eVects the musicians and moreover, what factors in their work envi- The study was designed to comprise all symphony orches- ronment that might contribute to the experience of stress. tra musicians from the seven professional Danish sym- Several studies among symphony orchestra musicians have phony orchestras; as one orchestra did not participate for found that social relations in particular can be an important administrative reasons, the study population was 441 musi- stressor (Parasuraman and Purohit 2000; Liljeholm Johans- cians from six orchestras. The response rate was 78%, 342 son and Theorell 2003). Among 12 Swedish classical respondents, of which 39.2% were women and 60.8% were orchestras, work content and social support were among the men. The distribution of gender did not diVer signiWcantly most important work-quality factors related to health between the respondents and the non-respondents. The aspects, and the higher social support, the fewer symptoms, overall demographic data are presented in Table 1. The and more well-being were reported (Liljeholm Johansson women were signiWcantly younger than men (p < 0.001) and Theorell 2003). Another Swedish study found that and constituted a large proportion of the high string player symphony orchestra musicians, compared with Wve other group and a low proportion of the low string players and occupations, reported high psychological demands and low brass players. The distribution of gender and age did not decision authority (Theorell and Ahlberg-Hulten 1990) diVer signiWcantly between the orchestras. A higher propor- which according to the demand-control model is a known tion of men had a secondary job. Similarly, more men lived combination of factors to increase the risk of ill health with a partner, and more females had received a candidate/ (Karasek 1979). master level from the conservatorium. 123 Int Arch Occup Environ Health

Table 1 DiVerences in sociode- Sociodemographics Male (n = 208) Female (n = 134) p mographic and occupational data in Danish symphony Individual orchestra musicians stratiWed by gender Age (year), mean (SD) 46.8 (11.3) 41.8 (10.6) <0.001 Age, range 35.5–58.1 31.2–52.4 Orchestra Orchestra 1, n (%) 28 (13.5) 21 (15.7) NS Orchestra 2, n (%) 45 (21.6) 24 (17.9) Orchestra 3, n (%) 17 (8.2) 18 (13.4) Orchestra 4, n (%) 32 (15.4) 26 (19.4) Orchestra 5, n (%) 32 (15.4) 18 (13.4) Orchestra 6, n (%) 54 (26.0) 27 (20.1) In total, n (%) 208 (100.0) 134 (100.0) Instrument group High string players 63 (23.2) 86 (39.1) <0.001 1. violinist, n (%) 32 (11.8) 33 (15.0) <0.05 2. violinist, n (%) 12 (4.4) 30 (13.6) <0.001 Viola players, n (%) 19 (7.0) 23 (10.5) <0.05 Low string players, n (%) 43 (15.9) 16 (7.3) <0.05 Woodwind players, n (%) 41 (15.1) 21 (9.5) NS Brass players, n (%) 46 (17.0) 7 (3.2) <0.001 Players of other instruments, n (%) 15 (5.5) 4 (1.8) NS In total, n (%) 208 (100.0) 134 (100.0) Work Number of weekly playing hours, mean (SD) 29.9 (11.0) 32.7 (9.7) NS Had secondary jobs (yes), n % 141 (67.8) 70 (52.2) <0.01 Family Living with a partner (yes), n (%) 170 (81.7) 91 (67.9) <0.001 Children at home (yes), n (%) 97 (46.6) 70 (52.2) NS Educational level Prevalence in number (n) and Conservatory of Music, n (%) 194 (93.3) 127 (94.8) NS percent (%) or mean, standard deviation (SD), and range are Candidate level or equal, n (%) 144 (69.2) 107 (79.9) <0.05 presented. Chi-square test for Bachelor level or equal, n (%) 17 (8.2) 7 (5.2) NS categorical variables (n >5), Educational level unspeciWed, n (%) 33 (15.9) 13 (9.7) NS Fischers exact-test (n <5) and Other education than 14 (6.7) 7 (5.2) NS student’s t test for continuous conservatory of music, n (%) variables

Reference population Questionnaire

A random sample of the Danish work force of 3,517 active The study was designed as a part of a cross-sectional survey workers aged 20–60 years reported in 2004–2005 on their of the work environment and health in Danish symphony psychosocial work environment and health by a postal musicians conducted during 2007–2008 (Paarup et al. questionnaire (National Research Centre for the Work 2011). The dimensions of psychosocial work environment Environment 2005). The questions were to a large extent and stress symptoms were investigated using the Copenha- similar to the questions that were later used for the sym- gen Psychosocial questionnaire (COPSOQII) developed by phony orchestra musicians in 2007–2008 for the present the National Institute of Occupational Health in Denmark study. The overall mean value across all professions for (Pejtersen et al. 2010). The questionnaire has been vali- each gender served as the reference for the mean value of dated based on factor analysis and the Cronbach’s alpha for the symphony orchestra musicians. internal consistency of categorizing items and computing 123 Int Arch Occup Environ Health scales (Bjorner and Pejtersen 2010). Four dimensions cov- Stress symptoms ered psychosocial work environment: work demands, work organization and job content, interpersonal relation and lead- Three scales were used: emotional-, cognitive-, and somatic ership, and work-individual interface, while stress symptoms stress symptoms. Each scale included four items each with covered emotional-, cognitive- and somatic stress. A detailed four response categories: (1) Whole time, (2) Large part of description of the items and scales and a guideline for calcu- time, (3) Small part of time, and (4) No time. The response lating these is available (Pejtersen et al. 2010). categories were given the scores (1) 100, (2) 66.6, (3) 33.3, and (4) 0, respectively, and the scale score was the average Sociodemographics and occupational variables of the included items. Individual scale scores were only calculated if at least two of the items in a scale were Wlled in Factors included in the study with a probable inXuence on and were otherwise omitted. Missing values constituted less the perceived psychosocial work environment and stress than 2% of the respondents on the stress scales. involved demographics described in terms of gender (Stansfeld et al. 1998), age (Jorm et al. 2005), living with a Statistical analyses partner (Mausner-Dorsch and Eaton 2000; Hope et al. 1999), one or more children living at home (Pugliesi 1999), Comparison of continuous variables was done using t test and educational level (Landsbergis et al. 2003). Occupa- or ANOVA (analysis of variance) and for categorical tional information included terms of orchestra membership, variables Chi-square test. For comparison of male with weekly playing hours, and having secondary jobs. Instru- female musicians, and comparison of musicians with ment played was categorized in the instrument groups: 1st Danish workers, additional criteria of an absolute diVer- violinists, 2nd violinists, other string players (viola, cello, ence of at least Wve points in the mean values were set to double bass), and other instruments (oboe, clarinet, bas- pinpoint only diVerences of clinical relevance (Pejtersen soon, Xute, trombone, trumpet, tuba, horn, piano, timpani, et al. 2010; Nübling 2006). Additionally was tested percussion, and harp). whether the diVerence was larger among females than among males (gender*workgroup interaction). Unpaired Psychosocial work environment t tests were used supplied with additional criteria of an absolute diVerence of at least Wve points in the mean val- Sixteen of 30 scales and one separate item from the original ues, again in order to pinpoint only diVerences of clinical COPSOQ questionnaire were used to describe the psycho- relevance. Multiple linear regression analysis was social work environment. These scales constituted four employed to measure the impact of instrument group on overall dimensions. (1) Demands at work: quantitative the scores of psychosocial work environment and stress demands, work pace, emotional demands, demands for hid- symptoms. Independent variables included: instrument ing emotions, cognitive demands, insecurity at work and group (1st violinist, 2nd violinists, other string players, role conXicts. (2) Work organization and job content: inXu- and other instruments), gender, orchestra (1 through 6), ence at work, possibilities for development, meaning of age, children living at home (yes/no), having a partner work, and commitment to the workplace. (3) Interpersonal (yes/no), and having a secondary job (yes/no). Dependent relation and leadership: role clarity, role conXicts, predict- variables included: COPSOQ-scales of work demands, ability, social support, sense of community and reward. (4) work organization and job content, interpersonal relations Work-individual interface: insecurity at work and job satis- and leadership, work-individual interface variables of faction. Each scale was based on 2–5 items except cogni- overall scales and scales of stress symptoms emotional-, tive demands only including 1 item and job satisfaction cognitive-, and somatic stress. In addition, multiple including 6 items. Each item had Wve response categories: regression analysis was employed to measure the eVect of (1) Very often, (2) Rather often, (3) Sometimes, (4) Rather the psychosocial variables on the three stress scales. Inde- seldom, and (5) Very seldom, the categories were given the pendent variables included psychosocial scales (work scores (1) 100, (2) 75, (3) 50, (4) 25, and (5) 0, respectively. demands, work organization and job content, interper- The scale value was calculated as an average score of the sonal relations and leadership, work-individual interface), items included. High work demand thus implies unfavored gender, age, orchestra, and instrument groups. Dependent condition while high values in the other scales are favor- variables included stress scales: emotional-, cognitive-, able. An individual scale score was only calculated if at and somatic stress. CoeYcients () and their 95% conW- least half of the items were Wlled in, otherwise they were dence intervals were calculated. The probability value for omitted. Missing values only aroused for the majority of evaluating statistical signiWcance was p · 0.05. All calcu- scales for less than 2% of the respondents although for less lations were done using STATA version 10.1 (Stata than 5% for social support. Corp., College Station, TX, USA). 123 Int Arch Occup Environ Health

Table 2 Psychosocial work environment and stress in Danish symphony orchestra musicians stratiWed by gender Danish symphony orchestra musicians df =340

Male (n = 208) Female (n =134) p DiVerence

Mean § (SD) Mean § (SD)

Psychosocial work environment factors Work demands 54.9 (9.7) 59.7 (8.9) <0.001 ¡4.8 Quantitative demands 41.6 (10.4) 44.6 (10.6) <0.01 ¡3.1 Work pace 57.7 (18.1) 66.0 (15.9) <0.001 ¡8.3 Emotional demands 54.1 (17.0) 59.5 (15.2) <0.01 ¡5.4 Demands for hiding emotions 48.7 (18.0) 54.3 (19.3) <0.01 ¡5.6 Cognitive demands 72.5 (22.2) 75.2 (21.2) NS ¡2.7 Work organization and job content 55.9 (13.0) 57.0 (12.0) NS ¡1.1 InXuence at work 27.8 (19.7) 25.4 (19.7) NS 2.4 Possibilities for development 60.9 (17.2) 63.1 (15.5) NS ¡2.2 Meaning of work 70.0 (67.7) 70.9 (68.2) NS ¡0.8 Commitment to the workplace 65.2 (15.8) 69.1 (14.2) <0.05 ¡3.9 Interpersonal relation and leadership 59.6 (11.0) 61.5 (9.9) NS ¡1.9 Role clarity 72.9 (15.7) 74.1 (13.8) NS ¡1.3 Role conXicts 46.6 (16.7) 46.2 (16.2) NS 0.4 Predictability 53.8 (16.1) 56.0 (16.7) NS ¡2.2 Social support 48.5 (20.9) 52.2 (19.5) NS ¡3.7 Sense of community 69.7 (20.1) 71.8 (15.6) NS ¡2.1 Reward 61.9 (14.7) 62.8 (15.3) NS ¡0.8 Work-individual interface: 66.3 (10.1) 67.4 (10.0) NS ¡1.1 Insecurity at work 24.2 (15.5) 25.6 (16.9) NS ¡1.4 Job satisfaction 56.6 (10.5) 59.7 (10.0) <0.01 ¡3.1 Stress Emotional stress 26.5 (15.3) 35.1 (16.9) <0.001 ¡8.6 Cognitive stress 17.0 (14.2) 22.0 (17.1) <0.01 ¡5.0 Somatic stress 18.1 (13.9) 27.4 (14.6) <0.001 ¡9.3 The table shows the mean values and standard deviation (SD) for each variable. DiVerences are calculated and t test is performed, and p values represent signiWcant diVerences in mean scores between genders. n indicates number; level of signiWcance p < 0.05; NS denotes non-signiWcant

Ethics reported higher commitment to the workplace and higher job satisfaction than their male colleagues. These Wndings The study was approved by The Danish Data Protection were, however, underneath the threshold of relevant diVer- Agency before the data collection in 2007 and 2008. The ence. In addition, female musicians reported on higher lev- musicians were informed about the study by written and els of stress symptoms on all the three scales. oral information. Consent was obtained by participation. The diVerences between instrument groups are shown in Table 3. 1st violinists reported higher work demands than 2nd violinists. However, none of these associations were Results signiWcant. Moreover, 1st violinists reported signiWcantly more emotional stress than the 2nd violinists. Compared to Symphony orchestra musicians other string players, 2nd violinists reported lower meaning of work. Other instrument groups reported a signiWcant Table 2 shows gender diVerences within the Danish sym- lower work pace compared with 2nd violinists. According phony orchestras showing that female musicians reported to the resource dimensions work organization and job content, on signiWcantly higher work demands. Females also interpersonal relation and leadership, and work–individual 123 Int Arch Occup Environ Health

Table 3 Mutiple regression analysis of psychosocial work environment and stress symptoms stratiWed by instrument group Instrument groups

2. violin 1. violin Other string players Other instruments

CoeV. (95% CI) CoeV. (95% CI) CoeV. (95% CI)

Psychosocial work environment factors Work demands Reference 3.67 (¡0.23 to 7.58) ¡2.09 (¡5.72 to 1.53) ¡1.40 (¡4.95 to 2.15) Quantitative demands Reference 2.02 (¡2.30 to 3.39) ¡0.61 (¡4.60 to 3.39) ¡0.71 (¡4.63 to 3.20) Work pace Reference 6.10 (¡0.82 to 13.02) ¡4.40 (¡10.79 to 1.99) ¡9.58 (¡15.84 to ¡3.31) Emotional demands Reference 3.87 (¡2.72 to 10.46) ¡3.50 (¡9.57 to 2.58) 2.27 (¡3.68 to 8.22) Demands for hiding emotions Reference 1.53 (¡6.18 to 9.25) ¡5.39 (¡12.52 to 1.73) ¡2.75 (¡9.72 to 4.22) Cognitive demands Reference 1.37 (¡7.68 to 10.42) ¡0.49 (¡8.86 to 7.89) ¡0.42 (¡7.68 to 10.42) Work organization and job content Reference 1.72 (¡3.38 to 6.81) 3.16 (¡1.57 to 7.89) 6.07 (1.44–10.70) InXuence at work Reference ¡0.78 (¡8.27 to 6.71) 3.76 (¡3.16 to 10.68) 11.86 (5.08–18.64) Possibilities for development Reference 3.76 (¡2.95 to 10.47) 5.28 (¡0.94 to 11.49) 8.13 (2.05–14.21) Meaning of work Reference 0.04 (¡0.01 to 0.09) 0.06 (0.01–0.10) 0.06 (0.01–0.10) Commitment to the workplace Reference ¡0.03 (¡6.30 to 6.24) ¡1.37 (¡7.16 to 4.42) ¡0.47 (¡6.14 to 5.20) Interpersonal relation and leadership Reference ¡0.34 (¡4.69 to 4.00) 3.43 (¡0.53 to 7.38) 1.26 (¡2.58 to 5.09) Role clarity Reference 4.00 (¡2.05 to 10.05) 4.02 (¡1.56 to 9.61) 1.59 (¡3.88 to 7.05) Role conXicts Reference 3.83 (¡2.80 to 10.47) ¡0.87 (¡6.98 to 5.25) 1.54 (¡4.45 to 7.53) Predictability Reference ¡0.42 (¡6.43 to 5.59) ¡0.32 (¡5.87 to 5.23) ¡3.80 (¡9.24 to 1.63) Social support Reference ¡3.08 (¡11.44 to 5.27) 5.71 (¡1.90 to 13.32) 8.95 (1.56–16.34) Sense of community Reference 1.47 (¡6.09 to 9.02) 4.80 (¡2.18 to 11.78) ¡1.24 (¡8.08 to 5.60) Reward Reference 2.21 (¡3.64 to 8.06) 3.14 (¡2.26 to 8.55) 4.25 (¡1.03 to 9.54) Work-individual interface Reference ¡2.31 (¡6.34 to 1.71) ¡0.49 (¡4.18 to 3.20) ¡3.70 (¡7.32 to ¡0.07) Insecurity at work Reference 0.48 (¡5.90 to 6.87) ¡2.34 (¡8.23 to 3.55) 2.48 (¡3.29 to 8.25) Job satisfaction Reference ¡3.04 (¡7.27 to 1.19) ¡2.13 (¡6.00 to 1.74) ¡3.77 (¡7.58 to 0.04) Stress Emotional stress Reference 6.83 (0.10–13.57) 1.53 (¡4.71 to 7.76) ¡4.30 (¡1.81 to 10.41) Cognitive stress Reference 2.09 (¡4.45 to 8.63) 1.86 (¡4.18 to 7.90) 1.92 (¡4.00 to 7.84) Somatic stress Reference 4.62 (¡1.40 to 10.65) ¡2.07 (¡7.65 to 3.50) ¡0.28 (¡5.75 to 5.18) Adjusted for gender, age, orchestra, children living at home, having a partner and having a secondary job. Bold denotes statitical signiWcance at p <0.05

interface, a tendency toward 2nd violinist reported lower Symphony orchestra musicians versus Danish workers scores than other instrument groups was found. However, the associations for these Wndings were insigniWcant. Table 5 shows that compared with Danish male and female workers, the symphony orchestra musicians reported more Associations between psychosocial work environment than 5% higher mean scores on work demand variables on and stress two and three out of Wve work demands factors, respectively, and lower scores on variables of work organization Table 4 shows the relation between psychosocial work and job content, interpersonal relations and leadership, and environment and stress symptoms adjusted for individual work–individual interface. The most prominent diVerences determinants. Both the overall scale of work demands and were the higher scores in musicians on emotional demands the sub-scales were positively correlated to three scales of and cognitive demands, and the lower scores on inXuence at stress symptoms. The remaining overall scales and sub- work, social support, sense of community, and job satisfac- scales were negatively correlated to stress symptoms and tion; still musicians reported higher scores on commitment associations were strong. The strongest associations were to the workplace compared with the Danish workers. Male found in the category of work demands and interpersonal musicians were diVerent from the male Danish workers in relations and leadership. relation to reporting lower possibilities for development. 123 Int Arch Occup Environ Health

Table 4 Mutiple linear regression analysis of psychosocial work environment and stress symptoms Emotional stress Cognitive stress Somatic stress

CoeV. (95% CI) CoeV. (95% CI) CoeV. (95% CI)

Psychosocial work environment factors Work demands 0.49 (0.31–0.68) 0.25 (0.06–0.45) 0.45 (0.29–0.62) Quantitative demands 0.14 (¡0.03 to 0.30) 0.19 (0.02–0.36) 0.12 (¡0.03 to 0.27) Work pace 0.06 (¡0.04 to 0.17) ¡0.01 (¡0.12 to 0.09) 0.13 (0.03–0.22) Emotional demands 0.34 (0.23–0.45) 0.19 (0.08–0.30) 0.27 (0.18–0.37) Demands for hiding emotions 0.19 (0.09–0.28) 0.17 (0.08–0.26) 0.17 (0.09–0.25) Cognitive demands 0.07 (¡0.01 to 0.15) ¡0.04 (¡0.12 to 0.04) 0.05 (¡0.03 to 0.12) Work organization and job content ¡0.19 (¡0.33 to ¡0.04) ¡0.22 (¡0.36 to ¡0.08) ¡0.14 (¡0.27 to ¡0.02) InXuence at work ¡0.06 (¡0.15 to 0.04) ¡0.07 (¡0.16 to 0.03) ¡0.08 (¡0.17 to 0.00) Possibilities for development ¡0.14 (¡0.25 to ¡0.03) ¡0.13 (¡0.23 to ¡0.02) ¡0.09 (¡0.19 to 0.00) Meaning of work ¡0.12 (¡0.22 to ¡0.01) ¡0.16 (¡0.27 to ¡0.05) ¡0.07 (¡0.17 to 0.02) Commitment to the workplace ¡0.10 (¡0.22 to 0.01) ¡0.13 (¡0.25 to ¡0.02) ¡0.05 (¡0.16 to 0.05) Interpersonal relation and leadership ¡0.42 (¡0.60 to ¡0.25) ¡0.34 (¡0.53 to ¡0.16) ¡0.25 (¡0.41 to ¡0.09) Role clarity ¡0.12 (¡0.24 to ¡0.01) ¡0.19 (¡0.31 to ¡0.07) ¡0.10 (¡0.21 to 0.00) Role conXicts ¡0.28 (¡0.38 to ¡0.17) ¡0.24 (¡0.34 to ¡0.13) ¡0.19 (¡0.29 to ¡0.09) Predictability ¡0.09 (¡0.21 to 0.03) ¡0.02 (¡0.14 to 0.10) ¡0.04 (¡0.15 to 0.06) Social support ¡0.07 (¡0.16 to 0.02) ¡0.05 (¡0.15 to 0.04) ¡0.05 (¡0.13 to 0.03) Sense of community ¡0.16 (¡0.25 to ¡0.06) ¡0.12 (¡0.21 to ¡0.02) ¡0.09 (¡0.18 to ¡0.01) Reward ¡0.24 (¡0.36 to ¡0.12) ¡0.25 (¡0.37 to ¡0.13) ¡0.13 (¡0.24 to ¡0.02) Work-individual interface ¡0.39 (¡0.57 to ¡0.21) ¡0.42 (¡0.60 to ¡0.24) ¡0.38 (¡0.54 to ¡0.27) Insecurity at work ¡0.20 (¡0.31 to ¡0.09) ¡0.23 (¡0.34 to ¡0.12) ¡0.22 (¡0.32 to ¡0.12) Job satisfaction ¡0.27 (¡0.44 to ¡0.09) ¡0.01 (¡0.02 to ¡0.01) ¡0.01 (¡0.01 to ¡0.01) Adjusted for gender, age, orchestra, instrument groups, children living at home, having a partner and having a secondary job. Bold denotes statitical signiWcance at p <0.05

The female musicians diVered from female workers in rela- negatively than the Danish work force in both genders. tion to work pace. According to stress symptoms, the Dan- Within the group of musicians, females reported their psy- ish symphony orchestra musicians scored higher on stress chosocial work environment and stress symptoms more symptoms than the Danish workers most evident in emo- negatively than the male musicians. Our Wndings indicate tional- and somatic stress for the female musicians. The that 2nd violinists experienced higher work pace and lower interaction term gives an indication whether diVerence resources than other instrument groups, whereas 1st violin- between musicians and the Danish workforce is more pro- ists perceived higher emotional stress compared with 2nd nounced in females than in males. This showed a signiW- violinists. In general, adverse psychosocial work environ- cantly larger diVerence in cognitive demands and work ment was strongly and positively associated with stress pace, and emotional stress in females than in males, while symptoms. the diVerence in emotional demands was lower in females Regarding the Wndings of gender diVerences, female due to a higher level in the female Danish workforce. musicians reported higher work demands and in particular higher stress symptoms compared with their male colleagues. Similar diVerences have been reported by others Discussion (Liljeholm Johansson and Theorell 2003). On the other hand, female musicians reported higher commitment to the Key Wndings work place and higher job satisfaction which are known strong motivational factors for choice of job and important The Wndings of this national study of psychosocial work factors of wellbeing. A part of this diVerence was inherited environment and stress showed that Danish symphony by a general higher rate of symptoms in the female work orchestra musicians reported their psychosocial work force, but the test of interaction indicated that work as a environment and prevalence of stress symptoms more musician is more strenuous to women than men resulting in 123 Int Arch Occup Environ Health

Table 5 Psychosocial work environment and stress in Danish symphony orchestra musicians and Danish workers stratiWed by gender Danish symphony orchestra Danish workers DiVerences between musicians musicians and Danish workers

Male Female Male Female Male Female Female (n = 208) (n = 134) (n = 1,668) (n = 1,849) diVerence diVerence diVerence versus male Mean § (SD) Mean § (SD) Mean § (SD) Mean § (SD) diVerence

Psychosocial work environment factors Work demands Quantitative demands 41.6 (10.4) 44.6 (10.7) 40.1 (20.0) 40.2 (20.8) 1.5 4.4 2.9 Work pace 57.7 (18.1) 66.0 (16.0) 58.6 (18.5) 60.1 (19.6) ¡0.9 5.9 6.8 Emotional demands 54.1 (17.0) 59.5 (15.2) 34.2 (22.7) 46.5 (24.2) 19.9 13.0 ¡6.9 Demands for hiding emotions 48.7 (18.0) 54.3 (19.3) 47.2 (21.1) 53.7 (19.8) 1.5 0.6 ¡0.9 Cognitive demands 72.5 (22.3) 75.2 (21.2) 64.3 (18.7) 46.5 (24.2) 8.2 28.7 20.5 Work organization and job content InXuence at work 27.8 (19.8) 25.4 (19.7) 51.8 (21.7) 47.8 (20.4) ¡24.0 ¡22.4 1.6 Possibilities for development 60.9 (17.2) 63.1 (15.6) 66.0 (17.6) 65.7 (17.5) ¡5.1 ¡2.6 2.5 Meaning of work 70.0 (67.7) 70.9 (68.3) 73.2 (15.7) 74.3 (15.7) ¡3.2 ¡3.4 ¡0.2 Commitment to the workplace 65.7 (15.9) 69.1 (14.2) 60.9 (19.8) 60.9 (20.8) 4.8 8.2 3.4 Interpersonal relation and leadership Role clarity 72.9 (15.7) 74.1 (13.8) 73.9 (16.3) 73.2 (16.3) ¡1.0 0.9 1.9 Role conXicts 46.6 (16.8) 46.2 (16.3) 42.5 (16.1) 41.4 (16.8) 4.1 4.8 0.7 Predictability 53.8 (16.1) 56.0 (16.7) 57.5 (21.3) 57.8 (20.4) ¡3.7 ¡1.8 1.9 Social support 48.5 (20.9) 52.2 (19.6) 55.0 (19.9) 59.3 (19.3) ¡6.5 ¡7.1 ¡0.6 Sense of community 69.7 (20.1) 71.8 (15.6) 79.3 (18.6) 78.1 (19.1) ¡9.6 ¡6.3 3.3 Reward 61.9 (14.8) 62.8 (15.3) 66.7 (19.1) 65.7 (20.5) ¡4.8 ¡2.9 1.9 Work-individual interface Insecurity at work 24.2 (15.6) 25.6 (17.0) 22.6 (19.5) 24.7 (21.8) 1.6 0.9 ¡0.7 Job satisfaction 56.6 (10.6) 59.7 (10.0) 65.3 (17.9) 65.1 (18.4) ¡8.7 ¡5.4 3.3 Stress Emotional stress 26.5 (15.3) 35.1 (17.0) 25.4 (17.0) 27.7 (18.1) 1.1 7.4 6.3 Cognitive stress 17.0 (14.3) 22.0 (17.1) 16.8 (15.1) 18.7 (16.0) 0.2 3.3 3.1 Somatic stress 18.1 (13.9) 27.4 (14.7) 14.6 (14.5) 20.6 (16.6) 3.5 6.8 3.3 DiVerences and interaction between gender and work are calculated and tested using unpaired t test. Values in bold represent diVerences of more than 5%, which all are signiWcant with p <0.01. n indicates number; mean and SD indicates standard deviation

higher perceived cognitive demands and emotional stress the present study also indicate that string players in general symptoms. experience lower inXuence and possibilities for develop- In relation to instrument groups, the 2nd violinists ment compared with other instrument groups. The Wndings reported lower resources compared with other instrument in this study distinguish between string players but do not, groups in accordance with the previously stated expecta- however, distinguish between the group of “other instru- tions. Liljeholm stated that social pressure to perform per- ments”. There may be important diVerences between string fectly is omnipresent during the carrier as professional players and wood wind players, brass players, percussion, symphony orchestra musician. Great emphasis is put on the and other but the groups were too small to make any dis- individual to contribute to a high artistic quality both on an tinctions. individual level but also in the orchestra as a whole Associations between psychosocial work environment (Johansson 2010). This may be part of the reason why the and stress showed that the musicians’ reporting of adverse 2nd violinists reported lower social support and reward work demands, work organization and job content, inter- compared with the other instrument groups. The Wndings in personal relations and leadership, and work–individual

123 Int Arch Occup Environ Health interface were signiWcantly associated with increasing in relation to work demands (Pejtersen et al. 2010). On the stress symptoms. Other studies have found speciWc working other hand, according to the dimensions work organization, conditions of musicians to be related to adverse psycholog- interpersonal relations, and work–individual interface, the ical outcomes such as interactions with the conductor, symphony orchestra musicians had similar scores to rehearsals, artistic integrity, solo performances, and perfor- slaughterhouse workers (Pejtersen et al. 2010). Therefore, mance anxiety (Parasuraman and Purohit 2000; Liljeholm the symphony orchestra musicians have a unique psychoso- Johansson and Theorell 2003; Salmon et al. 1995; Middles- cial work environment that cannot necessarily be compared tadt and Fishbein 1988). Social support has been identiWed with other occupational groups with similar educational as an important stressor in symphony orchestras in several levels. previous studies (Liljeholm Johansson and Theorell 2003; Johansson 2010; Theorell et al. 2007). However, in the Limitations present study, the Wndings on associations of social support and stress did not emphasize this factor in particular. Including six out of seven symphony orchestras in Den- Finally, the Wndings indicate that both male and female mark, with a response rate of 78%, the study population musicians experience more strenuous work conditions and was representative of the population of Danish symphony stress compared with Danish workers (National Research orchestra musicians according to sex, age, type of orches- Centre for the Work Environment 2005). A Norwegian tra, and geography, suggesting that the response bias is lim- study also found that musicians reported higher levels of ited. However, respondents and non-respondents may diVer gastrointestinal complaints and pseudoneurology com- thus non-respondents may have higher prevalence of stress plaints such as tiredness and mood changes compared with symptoms. If this is the case, our study may have underesti- a representative sample of the Norwegian population (Hal- mated the prevalence of stress symptoms. leland et al. 2010). The Wndings in this study were also con- The majority of the musicians in this study have been sistent with previous studies showing high work demands, playing music full time for decades and a major selection adverse job content, and high perception of stress being of can be expected to have taken place several years ago. concern (Theorell and Ahlberg-Hulten 1990; Liljeholm Based on this assumption, the risk of a healthy worker Johansson and Theorell 2003; Parasuraman and Purohit eVect is existent; however, this may potentially only 2000; Middlestadt and Fishbein 1988). In addition, low weaken the associations found. Likewise, a healthy worker inXuence was reported in the Danish musicians, which in eVect may be present in the population of the general work combination with high demands according to the demand- force which potentially may have increased the diVerence control-model, is known to cause adverse health eVects between musicians and Danish workers in our Wndings. such as cardiovascular diseases and therefore may be con- One main limitation of the study is inherent to the cross- sidered a serious risk (Karasek 1979). Moreover, social sectional design because both exposure and outcome are support is known to decrease the risk of high demands and measured at the same time representing the problem of low control (Karasek and Theorell 1990; Johnson and Hall common method bias (Kristensen 1996). Therefore, the 1988). However, our Wndings showed that both social sup- associations between perceived psychological work envi- port and the related variables sense of community and role ronment and stress symptoms found in this study may rep- conXicts were reported negatively in the musicians com- resent problems for Danish orchestra musicians but may pared with the Danish workers. In addition, our multivari- not be interpreted as causal relations. Moreover, assessment ate analysis showed that these variables individually were by questionnaire includes only selected variables represent- associated with the musicians’ experience of stress; how- ing a limited area of psychosocial work environment and ever, no combination of variables was analyzed. A positive stress. Stress reactions occur as an interaction between the Wnding related to the musicians reporting of higher commit- individual’s work situation (psychosocial and physical ment to the work place than the Danish workers. A possible work environment), private life situation, and personal fac- explanation of these Wndings might be that the musicians tors (genetics, gender, age, personality; Gadalla 2009). Sev- often commit themselves to work in the same symphony eral individual determinants were included in the analysis orchestra throughout their musical careers and thus develop to avoid confounding but due to the complexity of stress a strong attachment. potential, confounders may have been omitted. In relation to prevention, a comparison with speciWc A golden standard for measuring stress is not available; work groups is interesting. Pejtersen et al. have data for triangulation is suggested by comparing the self-reported many Danish professions from a national study that used measures of perceived stress with behavioral and physio- the same COPSOQ instrument. Compared to these Wndings, logical measures of stress (Kristensen 1996; European the group of Danish symphony orchestra musicians had Commission 2002). This was desirable, but not possible similar scores as academics in social- and human sciences in the given research context. The generic COPSOQ 123 Int Arch Occup Environ Health instrument was used in the study, and this theory-based through analysis of diVerential item functioning and diVerential instrument has been tested nationally as a reliable and valid item eVect. Scand J Public Health 38:90–105 Borritz M, Rugulies R, Bjorner JB, Villadsen E, Mikkelsen OA, Kris- method to assess psychosocial work environment and stress tensen TS (2006) Burnout among employees in human service reactions in a wide range of occupations (Pejtersen et al. work: design and baseline Wndings of the PUMA study. Scand J 2010; Arvidsson et al. 2008; Borritz et al. 2006). Informa- Public Health 34:49–58 tion bias can, however, occur in case of diVerent interpreta- European Commission (2002) Guidance on work-related stress: spice Y Y V of life kiss of death. Report, Luxembourg: O ce for O cial Pub- tions of the items within di erent groups, e.g., gender and lications of the European Communities instrument groups. However, our Wndings are very similar Fetter D (1993) Life in the orchestra. Md Med J 42:289–292 to Liljeholm and Theorell’s Wndings based on an adapted Gadalla TM (2009) Determinants, correlates and mediators of psycho- questionnaire for musicians (Liljeholm Johansson and logical distress: a longitudinal study. Soc Sci Med 68:2199–2205 Halleland HB, Harris A, Sørnes S, Murison R, Ursin H (2010) Subjec- Theorell 2003). tive health complaints, stress, and coping in orchestra muscians. Med Probl Perform Ar 24:58 Hamilton LH, Kellea JJ, Hamilton WG (1995) Personality and occupa- Conclusion tional stress in elite performers. Med Probl Perform Ar 10:86–89 Hope S, Rodgers B, Power C (1999) Marital status transitions and psychological distress: longitudinal evidence from a national popula- This study characterized the psychosocial work environ- tion Sample. Psychol Med 29:381–389 ment in the Danish symphony orchestras. Within the Johansson YL (2010) Psyksocial arbeitsmiljö I en yrkesgrupp med orchestras, female musicians and 2nd violinists seemed to krav på hög kvalitet—orkestrar inom konstmusik. Thesis, Karo- linska Instituttet—Department of Public Health Science, Sweden be of particular risk compared with the other instrumental Johnson JV, Hall EM (1988) Job strain, work place social support, and groups in aspects of work pace, work organization, and cardiovascular disease: a cross-sectional study of a random sam- content. Associations between the demanding psychosocial ple of the Swedish working population. Am J Public Health work environment and increasing stress symptoms could 78:1336–1342 Jorm AF, Windsor TD, Dear KJ, Christensen H, Rodgers B (2005) Age indicate a need for promotion and improvement of the group diVerences in psychological distress: the role of psychoso- musicians’ working conditions in order to prevent possible cial risk factors that vary with age. Psychol Med 35:1253–1263 adverse health eVects. Finally, both male and female sym- Karasek RA (1979) Job demands, job decision latitude and mental phony orchestra musicians reported a more demanding psy- strain: implications for job redesign. Adm Sci Q 24:285–307 Karasek R, Theorell T (1990) Healthy work. Basic Books, New York chosocial work environment exposure and a higher level of Kivimäki M, Jokinen M (1994) Job perceptions among symphony stress symptoms compared with what was earlier found orchestra musicians: a comparison with other occupational among the general Danish workforce. Our Wndings empha- groups. Med Probl Perform Ar 9:73–76 size the importance for future studies to focus more com- Kristensen TS (1996) Job stress and cardiovascular disease: a theoretic critical review. J Occup Health Psychol 1:246–260 prehensively on the work-related, private, and personal Laitinen H, Poulsen T (2008) Questionnaire investigation of musi- factors associated with stress reactions, and coping strate- cians’ use of hearing protectors, self reported hearing disorders, gies to deal with stressors in a symphony orchestra context and their experience of their working environment. Int J Audiol in order to get a deeper understanding of these relations. 47:160–168 Landsbergis PA, Schnall PL, Pickering TG, Warren K, Schwartz JE (2003) Lower socioeconomic status among men in relation to the Acknowledgments We would like to thank the participating sym- association between job strain and blood pressure. Scand J Work phony orchestras for taking part in the study. Thanks to the Danish Na- Env Hea 29:206–215 tional Research Centre for the Working Environment for providing Liljeholm Johansson Y (1996) Konstmusikers arbeitssituation-en data on the general work force. The study was funded by the Health explorativ studie med fokus på arbetstillfredsställelse och presta- Foundation, Denmark; The Carl Nielsen Academy of Music, Odense, tionsstress. SAMU, Sweden Denmark; The Faculty of Health Sciences and The Clinical Institute, Liljeholm Johansson Y, Theorell T (2003) Satisfaction with work task University of Southern Denmark, Odense, Denmark. quality correlates with employee health. Med Probl Perform Art 18:141–149 X Con ict of interest The authors declare that they have no competing Mausner-Dorsch H, Eaton WW (2000) Psychosocial work environ- interests. The funding organization had no role in the design or the con- ment and depression: epidemiologic assessment of the demand- duct of the study. control model. Am J Public Health 90:1765–1770 Middlestadt SE, Fishbein M (1988) Health and occupational correlates of perceived occupational stress in symphony orchestra musicians. J Occup Med 30:687–692 References National Research Centre for the Work Environment (2005) Psykisk arbejdsmiljø [psychosocial work environment, COPSOQ]. Avail- Arvidsson I, Axmon A, Skerfving S (2008) Follow-up study of muscu- able at: http://www.arbejdsmiljoforskning.dk/Nationale%20Data/ loskeletal disorders 20 months after the introduction of a mouse- 3DII.aspx. Accessed 10 Dec 2010 based computer system. Scand J Work Env Hea 34:374–380 Nübling M (2006) Measuring psychological stress and strain at Bjorner J, Pejtersen JH (2010) Evaluating construct validity of the sec- work—evaluation of the COPSOQ questionnaire in Germany. ond version of the copenhagen psychosocial questionnaire Psycho Soc Med 3:1860–5214

123 Int Arch Occup Environ Health

Ostwald PF, Baron BC, Byl NM, Wilson FR (1994) Performing arts Salmon P, Shook CP, Lombart K, Berenson G (1995) Performance medicine. West J Med 160:48–52 impairments, injuries, and stress hardiness in a sample of key- Paarup HM, Baelum J, Holm JCW, Maniche C, Wedderkopp N (2011) board and other instrumentalists. Med Probl Perform Ar 10:140– Prevalence and consequences of musculoskeletal symptoms in 146 symphony orchestra musicians vary by gender: a cross-sectional Stansfeld SA, Fuhrer R, Shipley MJ (1998) Types of social support as study. BMC Musculoskel Dis 12:223 predictors of psychiatric morbidity in a cohort of British civil ser- Parasuraman S, Purohit YS (2000) Distress and boredom among vants (Whitehall II study). Psychol Med 28:881–892 orchestra musicians: the two faces of stress. J Occup Health Psy- Sternbach D (1993) Addressing stress-related illness in professional chol 5:74–83 musicians. Md Med J 42:283–288 Pejtersen JH, Kristensen TS, Borg V and Bjorner JB (2010) The sec- Theorell T, Ahlberg-Hulten G (1990) A psychosocial and biomedical ond version of the Copenhagen psychosocial questionnaire. comparison between men in six contrasting service occupations. Scand J Public Health 38:8 Work Stress 4:51–63 Piperek M (1981) Stress and music: medical, psychological, sociolog- Theorell T, Liljeholm-Johansson Y, Björk H, Ericson M (2007) Saliva ical and legal strain factors in a symphony orkestra musician’s testosterone and heart rate variability in the professional sym- profession. Universitäts-Verlagsbuchhandlung Ges., Wien phony orchestra after “public faintings” of an orchestra member. Pugliesi K (1999) The consequences of emotional labor: eVects on work Psychoneuroendocrino 32:660–668 stress, job satisfaction, and well-being. Motiv Emot 23:125–154

123

View publication stats